Skip to content
All docs

AI Providers

OpenNapse is a bring your own key AI workspace. You choose the provider, keep control of the API key, and decide whether AI runs through hosted credits or your own infrastructure.

Supported providers

OpenNapse supports any provider with an OpenAI-compatible API. Configured providers include:

  • OpenAI — GPT-4o, GPT-4o-mini, and all models
  • Anthropic — Claude 3.5 Sonnet, Claude 3 Opus
  • Mistral — Mistral Large, Mistral Small
  • DeepSeek — DeepSeek V2, DeepSeek Coder
  • Groq — Fast inference with Llama and Mixtral
  • OpenRouter — Gateway to 200+ models
  • Ollama — Local models on your own machine

Bring Your Own Key

Every provider accepts your own API key. Keys stay in browser memory for the active session and are only sent to the provider you selected. On the hosted Cloud plan, the app keeps BYOK credentials client-side — the server does not persist them.

When you bring your own key, AI usage is free and does not count against your daily credit limit. Cloud Free gets 10 credits per day, Cloud Pro gets 500 — these cover the hosted AI inference path.

Provider comparison

  • OpenAI / Anthropic — strong general-purpose assistants
  • Groq / OpenRouter — flexible routing and speed
  • Ollama — fully local models for maximum privacy

Privacy FAQ

If you self-host and use Ollama, your data never leaves your network. AI runs entirely on your hardware, and no telemetry or usage data is collected.